Deblocking Filter Decision Segmentation for Memory Optimization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional image coding and decoding methods face inefficiencies in memory usage due to the way deblocking filtering is performed, leading to increased memory costs and computational complexity.
Innovation Solution
The method involves dividing image blocks into segments perpendicular to the block boundaries, allowing for individual judgment of deblocking filter application using sample pixels within each segment, thereby optimizing memory usage and reducing computational complexity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If deblocking filtering is performed using conventional methods on entire image blocks, then filtering quality is maintained, but memory usage increases and computational complexity increases
Solution Approach 1:
The patent divides image blocks into multiple segments (e.g., first segment, second segment, third segment) along block boundaries. Each segment is processed independently for deblocking filtering decisions, allowing the system to evaluate only relevant pixels within each segment rather than entire blocks. This segmentation reduces the quantity of pixels that need to be stored and processed, directly addressing the memory usage problem while maintaining filtering quality through localized decision-making.
2Measurement precision
If deblocking filtering is performed using conventional methods on entire image blocks, then filtering quality is maintained, but computational complexity increases
Solution Approach 1:
By segmenting image blocks into smaller regions and making independent deblocking decisions for each segment based on local pixel characteristics, the patent reduces the computational scope. Instead of evaluating entire blocks, the system only processes pixels within relevant segments, decreasing the number of computations required while preserving filtering quality through localized adaptive decisions.
Solution Approach 2:
The patent applies different deblocking filtering decisions to different segments based on their local characteristics. Each segment is evaluated independently using sample pixels from that specific segment, allowing the system to apply filtering only where needed and with appropriate intensity. This local quality approach reduces overall computational complexity by avoiding uniform processing of entire blocks, while maintaining filtering quality through context-aware local decisions.
3Quantity of substance
If segments are divided to reduce memory usage, then memory efficiency improves, but judgment accuracy for filter application may decrease
Solution Approach 1:
The patent carefully designs segment boundaries and sizes to balance memory efficiency with judgment accuracy. Each segment is defined to include sufficient sample pixels for accurate deblocking decisions while excluding unnecessary pixels to reduce memory usage. The segmentation strategy ensures that each segment contains enough information for reliable filtering judgment without requiring storage of entire block data.
Solution Approach 2:
The patent enhances judgment accuracy within each segment by using local pixel characteristics and relationships specific to that segment. The deblocking decision for each segment is made based on local sample pixels and their relationships, ensuring accurate judgment despite the reduced scope. This local quality approach compensates for the smaller segment size by focusing computational resources on relevant local features.
Data Source
AI summary
The present disclosure relates to deblocking filtering, which may be advantageously applied for block-wise encoding and decoding of images or video signals. In particular, the present disclosure relates to an improved memory management in an automated decision on whether to apply or skip deblocking filtering for a block and to selection of the deblocking filter. The decision is performed on the basis of a segmentation of blocks in such a manner that memory usage is optimized. Preferably, the selection of appropriate deblocking filters is improved so as to reduce computational expense.